Nvidia CEO has single-handedly tanked the Quantum Computing public market--and he was wrong anyway
Advanced Computing, AI/ML, Chips, and Extended Reality
Good morning,
President Biden has signed an executive order further limiting semiconductor exports, with the aim to help maintain a U.S. edge over China in AI. Much of the tech industry is pushing back saying that this executive order will actually do more harm both to the U.S. tech industry and to many of our closest allies.
Nvidia’s CEO, Jensen Huang, has single-handedly tanked the quantum computing public market. In comments last week, he stated that useable QC remained elusive. He argued that we won’t see useful QC within the next 15 years but will before 30 years. Eventually, he narrowed it down and said that we should have the tech within 20 years.
D-Wave Systems CEO Alan Baratz retorted that his company had commercial quantum computers that customers are using today to run their companies. Baratz explained that Huang seems only familiar with one approach to QC requiring error-correction, commonly known as gate-based quantum computing. A gate-based approach does require error correction, as it seeks to provide universal computation. By contrast, Baratz’ company uses a quantum annealing approach which doesn’t require the error correction. Quantum annealing is better suited for optimization problems and as a result has greater tolerance for errors.
So, Baratz is right that useful QC is available today, but really only using quantum annealing which is useful for some things but not others.
So, let’s return to the gate-based QC method. Is Huang right about the timeline there? Mark Zuckerberg thinks so. In an appearance on Joe Rogan’s podcast this week, Zuckerberg admitted that he’s not a QC expert but from what he knows, he agrees with Huang. But here’s the thing: predictions of future technology advancements is notoriously hard. We’ve talked about this challenge of predicting exponential systems a few months ago using the filling of Lake Michigan as an analogy.
Gate-based quantum computing likely is advancing like an exponential system. We at BoF are more optimistic than Huang, based on advancements we’ve seen over the past year, specifically. Temperatures at which quantum computers can operate is rising. Error correction is on the rise.
There’s no guarantee that we’ll see useful gate-based quantum computers in the next ten years, but we’re increasingly optimistic that we’ll see a significant disruptive development leading to ‘very useful’ quantum computers. After all, we are techno-optimists and we are building our future.
Alright, enough yapping. Let’s dig in to the news.
News Headlines:
Senate schedules 15 confirmation hearings this week including CIA and DoD chiefs (MT)
SCOTUS appears poised to uphold TikTok ban (BBC)
Final draft of Gaza truce deal [presented to sides after 'breakthrough', officials say (RT)
China builds specialized mobile piers for Taiwan invasion (TwN)
Zuckerberg says most companies need more 'masculine energy' (BBG)
Quantum Tech:
Nvidia's Huang: useful QC is ~20 years away, sends quantum stocks into nosedive (RT)
Zuck on Joe Rogan's podcast backs Huang's assessment (BZ)
CEO of D-Wave Systems throws BS flag over Huang's 'dead wrong' assertions (TS)
Analyst from DA Davidson dismisses Huang's comments as 'self-serving' (IBD)
HENSOLDT to support DLR on quantum radar optimization for defense applications (QI)
Quantum computers get automatic error correction for the first time (NS)
Scientists create split-electrons, unlocking new path for QC (STD)
Quantum algorithms could prompt faster solutions for complex simulations (QI)
China is building infrastructure for attosecond lasers, to 'see' ultrafast particles (QI)
Saudia Arabia lays out its strategic vision for the quantum era (QI)
AI / ML
Biden readies executive action on AI use for cyber defense, enhancing software security (POL)
Uranium prices surged to record high amid an AI energy demand frenzy (FT)
U.S. supports Musk argument in OpenAI lawsuit (RT)
AI startups drive VC funding resurgence, capturing record U.S. investment in 2024 (RT)
The U.K. seeks to build homegrown rival to OpenAI in bid to become world leader in AI (BBC)
OpenAI urges U.S. to prioritizes AI funding, regulation to stay ahead of China (RT)
Google maps the future of AI agents (VB)
HPE secures $1B AI server deal for Musk's XAI (BBG)
China continues advancing in AI, despite U.S. chip restrictions (T)
Study suggests AI tools decrease critical thinking skills (FBS)
Nanotech, Nanomaterials, Chips
U.S. tightens its grip on AI chip flows across the globe (RT)
Tech group urges U.S. to halt 'AI Diffusion' rule that would limit global access to AI chips (AP)
AP provides a deeper look at how the AI chip rule would work (AP)
TSMC begins producing 4-nanometer chips in Arizona (RT)
AI chips are improving faster than Moore's Law: Nvidia CEO (TC)
XR / AR / VR
Ukrainian elite forces use L3Harris-made AR night vision goggles in combat (DB)
In historic first, judge dons VR headset to experience crime (FBS)
France-based Aledia unveiled $200M microLED factory for augmented reality displays (VB)
U.S. Army soldiers train with VR to counter growing drone threats (AR)
Deal Flow:
VC
Infinite Reality, an AI-powered immersive tech and XR startup, raised a $3B round at a $12.25B valuation (PU)
Anthropic is set to raise $2B at a $60B valuation, in a round led by Lightspeed (RT)
Ndustrial raised a $19.5M Series B round led by ABB and GS Energy to scale its energy intensity optimization AI platform (PRN)
Paragraf collaboration with University of Birmingham secures $4.2M from Innovate U.K. and UKRI Future Leaders Fellowship to scale graphene for quantum techs (QI)
PE / M&A / Exits
SoftBank and its majoity-owned chipmaker Arm are exploring an acquisition of chip designer Ampere Computing, which was last valued at $8B in 2021 (BBG)
Opportunities:
The Air Force Academy is seeking sources that may be able to provide a dilution refrigerator for its quantum computing research (SAM)
Time is running out to submit research ideas to the Army's Research Lab for the Army AI Innovation Institute BAA covering command and control and large pre-trained models (SAM)
The NIST is conducting market research to identify potential vendors that can assist in developing evaluations and benchmarks of AI models' relevant cyber capabilities and risks (SAM)
Editor’s Picks:
Foreign Affairs has run a n excellent piece on the race to lead the quantum future and the implications of winning or losing.
Rana Foroohar believes that AI is taking the U.S. in a strange new direction where growth is lifted but only with significant social disruption.
Lighter Side:
Keep Building,
BOF